ATLANTA—Vince Calhoun, Distinguished University Professor of Psychology at Georgia State University and director of the Center for Translational Research in Neuroimaging and Data Science, has received a five-year, $3.3 million grant from the National Institute on Aging to use advanced machine learning to identify and separate subtypes of Alzheimer’s disease.
Alzheimer’s disease is the sixth leading cause of death in the U.S., but unlike other major killers such as cancer and heart disease, scientists are unable to easily diagnose Alzheimer’s or effectively treat it. In part, this is because there has been little work focused on integrating imaging and genomic data in the quest to find biomarkers, measurable indicators of the disease or its level of severity.
“When patients develop cognitive impairment, we need to learn how to predict whether those patients will decline further, which may be an indication of Alzheimer’s disease, or maintain a mild level of impairment,” Calhoun said. “Alzheimer’s also does not always present in the same way in every patient. People might have similar symptoms, but the brain changes are different. But you can’t see that just by looking at the symptom profile.”
Calhoun said that’s one reason why there have been so many failed trials to find effective treatments for Alzheimer’s. Scientists may need to use different types of drugs to treat different subtypes of the disease.
He and Jingyu (Jean) Liu, associate professor of computer science and an expert in imaging genomics, will further develop a new multivariate machine learning framework called flexible subspace analysis, which Liu created to pull together and examine networks of structural imaging, functional imaging and genomic data from patients with Alzheimer’s and related disorders.
“It’s well known that there are genetic markers related to Alzheimer’s, such as APOE, which increases the likelihood of developing late-onset Alzheimer’s, but there’s been much less investigation into how combinations of sites in the genome may be contributing,” Calhoun said. “Using these approaches that we’ve developed, we’re going to look for patterns of genomic data that are linked with patterns of structural data and functional data — essentially a multimodal fingerprint for subtypes of Alzheimer’s disease.”
An abstract of the grant, 1RF1AG063153-01A1, is available here.
Distinguished University Professor
Calhoun is the founding director of the Center for Translational Research in Neuroimaging and Data Science (TReNDS), which is focused on improving our understanding of the human brain using advanced analytic approaches.